Method and system for controlling a hybrid drive system, optimizing fuel consumption and pollutant emissions, computer program product, hybrid drive system, and vehicle

Description

[0001] The present invention relates to the field of engine control and more particularly the control of a hybrid propulsion system of a vehicle, with the aim of reducing polluting emissions.

[0002] A hybrid vehicle is a vehicle comprising at least one electric machine and a thermal engine to provide traction for the vehicle.

[0003] Reducing nitrogen oxide (NOx) emissions is a major challenge in the development of engines, particularly diesel engines. Strict approval thresholds lead to the use of very expensive exhaust gas aftertreatment systems. In this context, diesel hybridization is economically attractive provided that it allows for the reduction of NOx emissions at the source. The addition of an electric motor provides a degree of freedom in the choice of operating points for the internal combustion engine. Advantages in terms of fuel consumption and NOx emissions can therefore be prioritized. The size, and therefore the cost of the aftertreatment, can then be reduced, thus offsetting the additional cost of hybridization. As a bonus, fuel consumption is also significantly reduced, particularly thanks to energy recovery, such as regenerative braking.

[0004] Energy monitoring is therefore a key element in the development of diesel hybrid propulsion systems. It is more complex than in the case of gasoline, where taking into account consumption and catalyst temperature is sufficient. For a diesel hybrid engine, one of the challenges is finding a compromise between NOx emissions and fuel consumption.

[0005] Furthermore, the temperature of the aftertreatment system must be taken into account in energy monitoring. Indeed, minimizing NOx emissions at the engine outlet is not enough because the efficiency of the aftertreatment system must be taken into account, which has a major impact on pollutant emissions released into the atmosphere. Indeed, depending on the temperature of the aftertreatment, its efficiency can go from nothing to nothing. It is therefore essential to optimize the activation of the aftertreatment and to maintain it at a sufficient temperature during driving. A strategy that simply minimizes emissions at the engine outlet without taking into account the aftertreatment is not necessarily interesting in terms of reducing NOx emissions at the exhaust because the gains obtained at the engine outlet can be offset (and exceeded) by a reduction in the efficiency of the aftertreatment

[0006] We can distinguish two main families of energy management laws for hybrid vehicles.

[0007] The first family uses heuristic techniques, based on the experience of their designer who sets arbitrary rules. These heuristic laws were quickly adopted by manufacturers due to their ease of implementation and their robustness. The following documents illustrate examples of heuristic strategies allowing a reduction of NOx emissions for diesel hybrid vehicles: D. Ambuhl, A. Sciarretta, C. Onder, L. Guzzella, S. Sterzing, K. Mann, D. Kraft, and M. Küsell, A causal operation strategy for hybrid electric vehicles based on optimal control theory. In Proceedings of the 4th Symposium on Hybrid Vehicles and Energy Management, 2007. N. Lindenkamp, C.-P. Stöber-Schmidt & P. Eilts, Strategies for Reducing NOx and Particulate Matter Emissions in Diesel Hybrid Electric Vehicles, SAE Paper n°2009-01-1305, 2009.

[0008] However, heuristic approaches have two major drawbacks: first, they do not guarantee the optimality of the proposed solution, second, they are specific to a given application and therefore require significant calibration work each time they have to be deployed on a new application.

[0009] Conversely, the second family concerns "model-based" control approaches that guarantee the quality of the solution obtained to the accuracy of the model and, once developed, they are easily reusable on various vehicle applications, since it is then sufficient to update the physical parameters that differ. Such "model-based" approaches, based on optimal control theory, have therefore been widely used to solve the energy supervision problem of hybrid vehicles. The following documents illustrate such methods: A. Sciarretta, L. Guzzella, "Control of hybrid electric vehicles. Optimal energy-management strategies", Control Systems Magazine, vol. 27, no. 2, April 2007, pp. 60-70. R. Cipollone, A. Sciarretta, "Analysis of the potential performance of a combined hybrid vehicle with optimal supervisory control", Proc. of the IEEE International Conference on Control Applications, Munich, Germany, October 4-6, 2006 (invited paper). J. Liu, H. Peng, "Control optimization for a power-split hybrid vehicle", in Proc. of the American Control Conference, 2006

[0010] Initially, most of these publications were limited to optimizing fuel consumption. However, this criterion is not sufficient and such strategies can lead to a significant increase in pollutant emissions, particularly NOx emissions. A method for taking into account engine exhaust emissions was proposed in patent application FR 2 982 824 (US 2013 / 0131956) and in the document: O. Grondin, L.Thibault, and C. Querel- Transient Torque Control of a Diesel Hybrid Powertrain for NOx limitation, Engine and Powertrain Control, Simulation and Modeling, Volume #3 2012

[0011] This approach, although experimentally validated, does not allow for the control of pollutant emissions at the exhaust (outlet of the aftertreatment system). In some cases, this approach can degrade pollutant emissions at the exhaust: to minimize NOx at the outlet of the heat engine, this type of strategy tends to reduce the load on the operating points of the heat engine, which reduces the enthalpy production at the exhaust and can prevent the aftertreatment system from reaching its activation temperature. This is very problematic since the only emissions that count are those released into the atmosphere at the exhaust, and not those directly at the outlet of the heat engine.

[0012] The consideration of post-treatment thermal energy within an energy supervision based on optimal control was presented in the document: A. Chasse, G. Corde, A. Del Mastro, and F. Perez, "Online optimal control of a parallel hybrid with after-treatment constraint integration," in Proceedings of the IEEE Vehicle Power and Propulsion Conference, 2010.

[0013] This command concerns a gasoline application. The minimization criterion used in this study takes into account fuel consumption, with exhaust thermal optimization being ensured by taking the post-treatment temperature into account as a state within the optimization problem. Such an approach is effective for quickly optimizing consumption and accelerating the activation of the post-treatment. However, it does not directly consider pollutant emissions, and consequently, it can lead to their increase, particularly on Diesel applications.

[0014] Patent application GB 2483371 A describes a method for controlling a hybrid propulsion system of a vehicle, making it possible in particular to reduce the vehicle's consumption as well as polluting emissions.

[0015] Patent application EP 2055585 A2 describes a method for controlling a hybrid vehicle taking into account the temperature of the electrical machines.

[0016] Document US 2013 / 184913 A1 discloses a method for optimizing the consumption and pollutant emissions of a hybrid vehicle, using a cost function.

[0017] The subject of the invention relates to a method for controlling a hybrid propulsion system of a vehicle, in which a control (torques and / or state of the powertrain) is defined which minimizes the consumption and pollutant emissions at the output of the post-treatment system. The control method is based on minimizing a cost function of a model of the propulsion system. Thus, the method according to the invention makes it possible to simultaneously minimize fuel consumption and pollutant emissions, by taking into account the efficiency of the post-treatment system. In addition, the control method according to the invention makes it possible, by means of the model, to integrate the physical phenomena implemented by the hybrid propulsion system. The method according to the invention

[0018] The invention relates to a control method according to claim 1, a computer program product according to claim 10, a hybrid propulsion system according to claim 11 and a vehicle according to claim 12. Brief presentation of the figures

[0019] Other characteristics and advantages of the method according to the invention will appear on reading the following description of non-limiting examples of embodiments, with reference to the figures appended and described below. There figure 1 schematically illustrates the steps of the method according to the invention. The figure 2 illustrates a first embodiment of the method according to the invention. The figure 3 illustrates a second embodiment of the method according to the invention. The figure 4 illustrates an example of a hybrid propulsion system. The Figure 5represents a curve comparing the measured post-treatment temperature to the estimated post-treatment temperature as a function of time. The figure 6 illustrates, for an example, comparative curves of the vehicle speed for two control methods according to the prior art, and for an embodiment according to the invention. The figure 7 illustrates, for the same example, comparative curves of the state of charge of the battery for two control methods according to the prior art, and for an embodiment according to the invention. The figure 8 illustrates, for the same example, comparative curves of the cumulative NOx emissions at the engine outlet for two control methods according to the prior art, and for an embodiment according to the invention. The figure 9illustrates, for the same example, comparative curves of the cumulative NOx emissions at the outlet of the post-treatment system for two control methods according to the prior art, and for an embodiment according to the invention. The figure 10 illustrates, for the same example, comparative curves of the temperature of the post-treatment system for two control methods according to the prior art, and for an embodiment according to the invention. The figure 11 illustrates, for the same example, comparative curves of the efficiency of the post-processing system for two control methods according to the prior art, and for an embodiment according to the invention. Detailed description of the invention

[0020] The method according to the invention makes it possible to reduce fuel consumption and NOx emissions at the outlet of the post-treatment system for a hybrid propulsion system.

[0021] According to the invention, the method allows the control of a hybrid propulsion system of a vehicle, in particular an automobile, comprising at least one electric machine and at least one thermal engine (Diesel or gasoline). The electric machine is powered by an electrical energy storage system. The term electrical energy storage system includes any means of storing electrical energy such as a battery, accumulator, pack, modules, super-capacitors, etc. In the remainder of the description, the term battery is used to designate any of the means of storing electrical energy. The hybrid propulsion system further comprises a drive train for coupling the thermal engine and the electric machine. This may be a series or parallel or mixed series / parallel drive train. The drive train may comprise reduction means, such as a gearbox, reducers, etc., coupling means, such as clutches... The hybrid propulsion system also includes a system for post-treatment of pollutant emissions (particularly NOx) from the thermal engine. The most common systems for reducing NOx are exhaust gas recirculation and selective catalytic reduction. In addition, a particulate filter can be used for hydrocarbons HC, carbon monoxide CO and fine particles.

[0022] For the method according to the invention, the following steps are carried out: acquisition of a torque setpoint from the propulsion system TPT sp;discretization of at least part of the set of commands admissible by the propulsion system, making it possible to reach the torque setpoint of the propulsion system; construction of a model of the propulsion system which links a cost function to a command of the propulsion system, the cost function being a function of the consumption of the propulsion system and the pollutant emissions at the output of the post-treatment system; determination of a command of the propulsion system by minimization of the cost function for the discretized admissible commands; and application of the command to the propulsion system.

[0023] According to the invention, the determined command can be a torque setpoint of the thermal engine T eng_sp and / or a torque setpoint from the electric machine T word_sp and / or a drivetrain control instruction ECC sp ,for example, a command instruction for the gear ratio of the drive train.

[0024] The torque setpoint of the propulsion system TPT sp corresponds to the driver's wheel torque demand.

[0025] The ordering method according to the invention is carried out online, in real time. Thus, it determines the order without knowing the vehicle's route in advance. Ratings

[0026] In the remainder of the description, the following notations are used: Born Engine speed [rpm] SOC Battery charge status [%] TPT sp Raw (unfiltered) driver wheel torque setpoint [Nm] TPT flt_sp Filtered driver wheel torque setpoint [Nm] v adm Vector of admissible orders [-] T eng Thermal engine torque [Nm] T eng_v Vector of admissible thermal engine torques [Nm] T eng_sp Thermal engine torque setpoint [Nm] T word Torque of the electric machine [Nm] T word_sp Electric machine torque setpoint [Nm] V veh Vehicle speed [km / h] ECC State of the kinematic chain at time t [-] ECC sp Drivetrain Status Instruction [-] ECC v Vector of admissible kinematic chain states [-] H v Hamiltonian Vector (Cost Function) [W equivalent] η AT Post-treatment system efficiency [-] T AT Aftertreatment system temperature [°C] T AT QS Steady-state post-treatment system temperature [°C] C name Nominal battery capacity (or equivalent) [C] OCV Battery open circuit voltage (or equivalent) [V] DCR Internal resistance of the battery (or equivalent) [Ω] λ Lagrange multiplier [-] u 1 Thermal engine torque control [Nm] u 2 Drivetrain Status Control [-] x Battery charge status [%] α Calibration variable [-] mf Fuel consumption of the thermal engine [kg / h] m NOx TP NOx emissions from the aftertreatment system [g / h] m NOx EO NOx emissions from the thermal engine [g / h] l Equivalent thermal inertia of the post-treatment system [W / K] k 1 Thermal resistance of exchanges with the exterior [W / K] k 2 Thermal resistance of exchanges with exhaust gases [W / K] Δt Time interval [s] R1 Reduction ratio of the reducer coupled to the electric machine [-] RBV Gearbox ratio coupled to the thermal engine [-] Pelec Power of the inverter feeding the electric machine [W] λ sp Calibration variable [-] K p Calibration variable [-]

[0027] The derivative with respect to time is indicated by a dot above the variable.

[0028] There figure 1 illustrates the different stages of the process according to the invention: acquisition of a torque setpoint from the propulsion system TPT sp ; DIS discretization of at least part of the set of admissible orders in adm. by the propulsion system, allowing the torque setpoint to be reached TPT spof the propulsion system; construction of a MOD model of the propulsion system which links a cost function H to a command of the propulsion system, the cost function H being a function of the consumption of the propulsion system and the pollutant emissions at the output of the post-treatment system; determination of a COM command of the propulsion system by minimizing the cost function for the admissible commands in adm. discretized. 1) Discretization

[0029] During this step, we discretize all admissible commands, allowing us to obtain the torque setpoint TPT sp of the hybrid propulsion system. Discretization consists of creating a mesh of all admissible control solutions. The pitch of this mesh can be chosen according to a compromise between the precision of the solution (fine mesh) and the acceleration of the calculation time (coarse mesh).

[0030] One discretization method that can be used is to create a regular mesh. This means that the mesh pitch, i.e. the distance between two elements, is constant. In this case, each element of the admissible control vector is obtained by the following equations: v adm i = T eng mini + ϵ ∗ i − 1

[0031] Where ε, the mesh pitch is simply obtained by fixing the number of mesh elements N (for example, we can choose N=10, which is an order of magnitude allowing a good compromise between precision and speed). For example: ϵ = T eng max − T eng min N − 1 with Tengmax the maximum permissible torque of the thermal engine allowing the torque setpoint to be reached, and Tengmin the minimum permissible torque of the thermal engine allowing the torque setpoint to be reached.

[0032] In this step, we therefore determine a vector of admissible orders in adm. . According to the invention, the vector of admissible commands in adm.can understand the vector of admissible thermal engine torques T eng_v .

[0033] Furthermore, the vector of admissible orders in adm. includes the vector of admissible kinematic chain states ECC v. 2) Model construction

[0034] In this step, a model of the hybrid propulsion system is constructed. The hybrid propulsion model is representative of the propulsion system's drivetrain. It can also take into account the battery's state of charge. The hybrid propulsion model links a cost function to a propulsion system command. The cost function is a function of the propulsion system's consumption and the pollutant emissions at the output of the aftertreatment system. It allows the calculation of the cost associated with each possible command, in terms of consumption and pollutant emissions (notably NOx). This cost is a calibratable compromise between exhaust pollutant emissions and fuel consumption.

[0035] The control method according to the invention applies to all hybrid architectures: series, parallel or mixed series / parallel. Depending on the architecture used, the equations modeling the hybrid powertrain are different, but the overall principle remains the same. In addition, the main added value of the invention is independent of the hybrid architecture considered, since it involves taking into account polluting exhaust emissions in addition to consumption. Consequently, the modeling of the hybrid powertrain is presented, in a non-limiting manner, in the case of a parallel hybrid propulsion system. The architecture considered is illustrated (in a non-limiting manner) on the figure 4The hybrid propulsion system comprises an ICE thermal engine, a Stop & Start generator SSG (electric machine that allows the automatic stopping and restarting of a thermal engine), a CL clutch, a GB gearbox, and an EM electric machine. The W wheels are coupled to the hybrid propulsion system by coupling means shown schematically.

[0036] For a parallel hybrid propulsion system, the wheel torque balance is written: TPT sp t = R 1 × T mot t + RBV ECC t × T eng t

[0037] We therefore have two degrees of freedom to carry out the driver's request. By convention, we choose here as the command the torque of the thermal engine u 1 = T a n g ( t ) and the state of the powertrain u 2 = ECC(t). We note that through the state of the kinematic chain we control the speed of the thermal engine No ( t ) = f BV ( u 2) ,where f BV characterizes the transmission reduction ratios.

[0038] These degrees of freedom are used to minimize a calibratable compromise (using the parameter α) between consumption mf fuel and pollutant emissions m NOxTP at the output of the post-processing system: J = ∫ t 0 tf f u 1 u 2 t dt f u 1 u 2 t = 1 − α × m ˙ f u 1 u 2 t + α × m ˙ NO x TP u 1 u 2 t

[0039] In addition, the dynamics of the battery charge state are taken into account. x(t) = SOC ( t ) . This state of charge is not completely free since the battery capacity is limited. At the optimization problem level, this amounts to adding a state constraint to the problem, so that the battery charge at the end tf of the request is identical to the battery charge state at the start t 0 of the request: SOC t f = SOC t 0

[0040] The cost function of the model can be given by a function of the type: H u 1 u 2 x t = f u 1 u 2 t + λ t × x ˙ u 1 u 2 x t Calculating f ( u 1 ,u 2 ,t )

[0041] To calculate f at each instant, we determine all the terms of the equation of f.

[0042] α is a calibration parameter that allows the compromise between fuel consumption and pollutant emissions to be adjusted. The calibration of the parameter α depends on the level of pollutant emissions of the engine in question. In general, a setting of α between 0.2 and 0.5 can be chosen, which ensures a significant reduction in pollutant emissions while maintaining attractive fuel consumption.

[0043] According to one embodiment of the invention, ṁ f = MAP ( No, Teng ) can be obtained by means of a fuel flow map (MAP), generally resulting from tests.

[0044] According to the invention, the NOx emissions at the exhaust outlet (from the post-treatment system) are modeled by an equation of the form: m ˙ NO x TP u 1 u 2 x t = m ˙ NO x EO u 1 u 2 x t × 1 − η AT T AT

[0045] NOx emissions from the engine ṁ NOx EO can either be calculated from a model or simply via a map from engine bench tests.

[0046] The characteristic of the efficiency of the aftertreatment system η AT as a function of the temperature T AT can be derived from characterization tests. It is also possible to take into account the influence of other variables, such as the exhaust gas flow rate.

[0047] The post-treatment temperature T AT can be estimated at each instant from the following equations: T AT t = T AT t − Δ t + Δ t × h 1 t + h 2 t I

[0048] Where the term h 1 corresponds to exchanges with the outside: h 1 t = k 1 × T 0 − T AT t − Δ t

[0049] And where the term h 2 corresponds to the release of enthalpy in the exhaust linked to combustion: h 2 t = k 2 × T AT QS u 1 t − Δ t , u 2 t − Δ t − T AT t − Δ t

[0050] The term T AT QS ( u 1 ( t - Δ t ), u 2 ( t - Δ t )) can be obtained by means of a mapping from tests, and corresponds to the temperature measured at the post-treatment level in steady state. The values of the parameters k1 and k2 can be determined from vehicle tests. This temperature model, although simplified for computational time constraints linked to integration into the energy management strategy, gives a correct representativeness, as illustrated in the figure 5. This figure compares the temperature of the aftertreatment system measured MES to this same temperature estimated EST using the previous equations, as a function of time and for a portion of the WLTC driving cycle (Worldwide Harmonized Test Procedure for Passenger Cars and Light Commercial Vehicles). In this figure, we observe that the model could be validated for different vehicle configurations (all thermal, Stop & Start and Full Hybrid: hybridization is total, the two engines ensure locomotion) and for the energy supervision strategy according to the invention (criterion for optimizing consumption and pollutants at the output of the aftertreatment system). Calculation of ẋ ( u 1 , u 2 , x, t )

[0051] The calculation of the dynamics of the system state which is the state of charge of the battery can be described in the following equations: x ˙ u 1 u 2 x = − I bat u 1 u 2 x C nom × 100

[0052] Using a model of the battery as an electric cell, the battery current can be expressed by an equation of the type: I bat = OCV x 2 × DCR x − OCV x 2 4 × DCR x 2 − P elec u 1 u 2 DCR x

[0053] With OCV and DCR respectively the no-load voltage and internal resistance of the battery depending on its state of charge, these characteristics can be derived from tests.

[0054] The calculation of the power P elec of the inverter supplying the electrical machine can be given by an equation of the form: P elec u 1 u 2 x t = f ME u 1 u 2

[0055] Where f ME is a mapping integrating the efficiency of the electrical machine and the inverter, resulting from tests dependent on its operating point, and therefore implicitly from the commands u 1 and u 2 . Calculating λ ( t )

[0056] The calculation of the Lagrange multiplier can be done via the following equation: λ t = λ SP + K p x SP − x t

[0057] Where λ SP and K p are calibrated to best contain the battery state of charge within its useful range and x SP being the average value of the battery state of charge.

[0058] Using these different equations, we can therefore estimate f ( u 1 , u 2 , t ) , ẋ ( u 1 , u 2 , x, t ) And λ ( t ). Thus, the cost function H of the hybrid propulsion model is fully determined. 3) Minimization of the cost function

[0059] In this step, the cost function H of the hybrid propulsion system model is minimized. The minimization is carried out on the admissible commands discretized in step 1). The discretization thus makes it possible to reduce the computation time required for the minimization.

[0060] This step consists of minimizing the vector of costs (Hamiltonians given by the equation of H) associated with each admissible order in order to determine which is the optimal order.

[0061] According to one embodiment of the invention, the minimization is implemented by means of the Pontryagin minimum principle.

[0062] According to one embodiment of the invention, the minimization step makes it possible to determine optimal torque setpoints for the thermal engine and / or the electric machine and / or an optimal control setpoint for the kinematic chain. 4) Application of the order

[0063] The invention makes it possible to determine the torque setpoints of the drive means of the hybrid propulsion system and / or a control setpoint of the drive train. By applying these setpoints to the thermal engine and / or to the electric machine and / or to the drive train, a reduction in pollutant emissions is obtained, and fuel consumption can also be limited.

[0064] The period of dynamic optimization is adapted to the physical phenomena involved at the engine level, in this case the production of polluting emissions.

[0065] The method according to the invention can be used for motor vehicles. However, it can be used in the field of road transport, the field of two-wheelers, the railway field, the naval field, the aeronautical field, the field of hovercraft, and the field of amphibious vehicles...

[0066] The method according to the invention is particularly suitable for "Full Hybrid" propulsion systems, but may also be suitable for "Stop & Start" or "Mild Hybrid" hybridization. A "Full Hybrid" hybridization corresponds to a system with a complete hybrid engine for which the electric motor(s) have sufficient power to provide propulsion on their own for a limited time. "Stop and Start" hybridization corresponds to a control system that stops the combustion engine when the vehicle is stationary in neutral and causes it to restart at the first request from the driver by means of a low-power electric machine. The "Mild Hybrid" type of propulsion system is equipped with a low-power electric machine and a braking energy recovery system, which provide additional power at low engine speeds or during a phase of strong acceleration.For a “Mild Hybrid” propulsion system, the electric machine is not capable of providing traction for a vehicle alone. Variants of realization

[0067] According to an implementation of the invention (which can be combined with all the variant embodiments described), the discretization can also be a function of the state of charge SOC of the vehicle and the speed of the vehicle. In the car.

[0068] According to a first embodiment of the invention, the torque setpoint of the hybrid propulsion system is filtered, the steps of the method being carried out for the filtered setpoint. The filtering may consist of a preventive anti-jerk filter, which filters the driver's torque request, so as to limit jerks.

[0069] According to a variant of this first embodiment of the invention, the control method determines the torque setpoints of the thermal engine T eng_spand the electric machine T mot_sp from the filtered propulsion system torque setpoint. The figure 2 illustrates the steps of the control method for this embodiment variant. According to this embodiment variant, the discretization is carried out from a filtered value TPT flt_sp of the torque setpoint of the hybrid propulsion system. In a non-limiting manner, the discretization is also a function of the state of charge SOC of the vehicle and the vehicle speed In the car. The discretization step makes it possible to determine the admissible thermal engine torques T eng_v . The modeling and minimization steps remain unchanged, compared to the embodiment described in connection with the figure 1 The minimization step makes it possible to determine the torque setpoints of the thermal engine. T eng_sp and the electric machine T mot_sp·

[0070] According to a second embodiment of the invention, the determined command corresponds to the state of the kinematic chain, for example to the control of the ratio of the gearbox of the kinematic chain. For this embodiment, the calculation principle (discretization, modeling, and minimization) is the same as for the optimization of the torque with the difference that the vector of admissible commands in adm. is not limited to the set of possible pairs T eng_v , but also contains all admissible kinematic chain states ECC v. Indeed, to be able to determine what the optimal kinematic chain state is, it is preferable to first determine what the optimal torque distribution is for each of the kinematic chain states to be compared. Indeed, it is the comparison of the costs of the optimal torque distributions for each admissible kinematic chain state that makes it possible to determine what the optimum is.

[0071] There figure 3 illustrates the steps of the method according to this second embodiment. The discretization is carried out from an unfiltered value TPT sp of the torque setpoint of the hybrid propulsion system. In a non-limiting manner, the discretization is also a function of the state of charge SOC of the vehicle and the vehicle speed In the car. The modeling and minimization steps remain unchanged, compared to the embodiment described in connection with the figure 1 (or the figure 2 ), apart from the fact that the vector of admissible orders in adm. further includes all admissible kinematic chain states ECC v. The minimization step makes it possible to determine a control instruction for said kinematic chain. ECC sp. which can be a gear ratio instruction from the drivetrain gearbox.

[0072] According to a third embodiment of the invention, a torque setpoint of the thermal engine is determined T eng_sp and / or a torque setpoint from the electric machine T mot_sp using the filtered propulsion system torque setpoint TPT flt_sp and steps 1) to 3) are repeated to determine a control setpoint for the kinematic chain ECC sp using the unfiltered torque setpoint TPT sp. and the commands are applied to the hybrid propulsion system. For this embodiment, steps 1) to 3) are therefore repeated twice: once with a filtered torque setpoint TPT flt_sp and once with a raw torque setpoint TPT sp(unfiltered). These two determinations can be implemented in parallel. The interest in separating the optimization of the torque from that of the powertrain state comes from the impact of the approval strategies which modify the torque demand at the driver's wheel. These filtering strategies, such as the preventive anti-kick filter, filter the driver's torque demand TPT sp depending on the current state of the drivetrain. The optimization of the torque distribution is carried out from this filtered torque setpoint TPT flt_sp . But, to choose the optimal kinematic chain state, it is better to take the raw torque setpoint TPT sp. The need to use different input signals justifies having to perform two optimizations in parallel.

[0073] According to an alternative embodiment, the repetitions of steps 1) to 3) can be implemented according to the alternative embodiments illustrated in the figures 2 and 3 .

[0074] The invention further relates to a computer program product downloadable from a communications network and / or recorded on a computer-readable medium and / or executable by a processor or a server. This program comprises program code instructions for implementing the method as described above, when the program is executed on a computer or a controller.

[0075] Furthermore, the invention relates to a hybrid propulsion system for a vehicle. The hybrid propulsion system comprises at least one electric machine powered by an electrical energy storage system, a heat engine, a drive train and a system for post-treatment of pollutant emissions (in particular NOx) from the heat engine. The hybrid propulsion system comprises control means for carrying out the following steps: acquisition of a torque setpoint from the propulsion system TPT sp; discretization of at least part of the set of commands admissible by the propulsion system, making it possible to reach the torque setpoint of the propulsion system; construction of a model of the propulsion system which links a cost function to a command of the propulsion system, the cost function being a function of the consumption of the propulsion system and the pollutant emissions at the output of the post-treatment system; determination of a command of the propulsion system by minimization of the cost function for the discretized admissible commands; and application of the command to the propulsion system.

[0076] The control means may be compatible with all the variants of the control method described above.

[0077] Furthermore, the invention relates to a vehicle comprising such a hybrid propulsion system. The vehicle according to the invention may be a motor vehicle. However, it may be any type of vehicle in the field of road transport, the field of two-wheelers, the railway field, the naval field, the aeronautical field, the field of hovercraft, and the field of amphibious vehicles... Comparative example

[0078] In order to demonstrate the advantages of the control method according to the invention. The control method according to the invention is compared with control methods according to the prior art. The examples proposed here are experimental validations of the results on the engine test bench.

[0079] Application cases: Utility vehicle, mass: 2700kg Diesel engine: 120kW Electric motor: 50kW

[0080] The control process was tested and compared to the engine test bench on the WLTC cycle which will be the official homologation cycle from the Euro7 standard.

[0081] The test equipment used is a high dynamic engine test bench with a gas analysis bay for measuring pollutant emissions.

[0082] The engine used has an extended EGR zone, standard in Euro6C. This is an important clarification because until now many energy monitoring projects reducing consumption and pollutant emissions have only been validated with Euro 5 engines. However, a Euro 5 diesel engine has two relatively different operating zones: in the rather reduced zone corresponding to the operating points of the driving cycle (NEDC), exhaust gas recirculation (EGR) is used. As a result, this zone has low levels of nitrogen oxides and degraded consumption. In the second zone, outside the cycle, the engine settings are optimized on a consumption criterion alone, and EGR is not used. Therefore, it is not surprising that energy monitoring manages to significantly vary the compromise between NOx emissions and fuel consumption.

[0083] For this example, the control method according to the invention is compared to a control method according to the prior art, for which only the consumption of the vehicle is optimized, and is compared to a control method according to the prior art, for which the consumption and the pollutant emissions at the engine outlet (before post-treatment) are optimized.

[0084] The experimental results are given in Table 1 and on the figures 6 to 11 The control method according to the invention is indicated INV, the method according to the prior art, for which only consumption is optimized, is noted AA1 and the method according to the prior art, for which consumption and emissions at the engine outlet are optimized, is noted AA2. figure 6 shows the vehicle speed curves V vehicle as a function of time for the three processes. The figure 7 concerns the battery state of charge SOC curves for the three processes. The figure 8concerns the accumulation of polluting emissions from the thermal engine m NOx EO for all three processes. The figure 9 illustrates the curves of the cumulative pollutant emissions at the outlet of the post-treatment system m NOx TP for the three processes compared. The figure 10 represents the temperature curves of the post-treatment system T AT for the three control processes compared. The figure 11 illustrates the curves of the efficiency of the post-treatment system the AT for all three control processes.

[0085] In Table 1 and the figures 6 to 11 , we observe that each control process allows us to obtain the same speed profile ( figure 6) than the control methods according to the prior art AA1 and AA2. In addition, it can be observed that each method makes it possible to minimize the associated parameter, since in each of the three cases the minimum value of the optimized associated parameter is obtained. Compared to the state of the art AA1, which corresponds to the strategy optimizing fuel consumption, it is observed that it is possible to reduce, by means of the control method according to the invention INV, the nitrogen oxide NOx emissions at the exhaust outlet significantly (55%) at the cost of a slight increase in consumption (2%). To do this, the control method according to the invention INV initially performs all-electric driving, illustrated by the drop in the battery state of charge on the figure 7 . Then, the thermal engine is switched on and used on points favoring the activation of the post-treatment, which turns out to be very fast, as illustrated in the figure 11, and effective as illustrated on the figure 10 . As a result, the cumulative NOx emissions from the exhaust remain low as illustrated in the figure 9 , while the cumulative NOx emissions at the engine outlet are close to those of the prior art AA1 process ( figure 8 ). Table 1 - Comparative example Fuel consumption [L / 100km] Nitrogen oxides from engine [mg / km] Nitrogen oxides exhaust [mg / km] AA1 8.25 466 101 AA2 8.57 249 72 INV 8.42 446 45

Claims

1. Method for controlling a hybrid propulsion system, comprising at least one electric machine, at least one combustion engine, at least one system for storing electric energy powering said electric machine, a kinematic chain for coupling said electric machine and said combustion engine, and a system for after-treatment of polluting emissions output from said combustion engine, wherein a torque setpoint TPTsp is obtained for said propulsion system, this comprising the following steps: a) the set of commands acceptable to said propulsion system is discretized, with a view to achieving said torque setpoint TPTSP of the propulsion system, by generating a mesh of the set of acceptable commands forming a vector of acceptable commands containing a vector of torques acceptable to the combustion engine; b) a model of said propulsion system is constructed, which model relates a cost function to a command applied to said propulsion system, said cost function being a function of the consumption of said propulsion system and of the polluting emissions output from said after-treatment system, said cost function of the model of the hybrid propulsion system being expressed by a formula of the following type H(u1, u2, x, t) = f (u1, u2, t) + A (t) × ẋ (u1, u2, x, t) with f(u1, u2, t) = (1 - α) × ṁf(u1, u2, t) + α × ṁNOxTP (u1, u2, t) and ṁNOxTP (u1, u2, x, t) = ṁNOxEO (u1, u2, x, t) × (1 - ηAT (TAT)) with u1 the torque command Teng applied to said combustion engine, u2 the command applied to said kinematic chain ECC, x the state of charge of said system for storing electric energy, mf the consumption of said combustion engine, mNOxTP the NOx emissions output from said after-treatment system, α a calibration variable, A the Lagrange multiplier, t time, and mNOxEO the polluting emissions output from said combustion engine, ηAT the efficiency of said after-treatment system, and TAT the temperature of said after-treatment system; c) a command to be given to said propulsion system is determined by minimizing said cost function of said model of the propulsion system in light of said discretized acceptable commands; and d) said determined command is applied to said hybrid propulsion system.

2. Method according to Claim 1, wherein said command is a torque setpoint Teng_sp of said combustion engine and / or a torque setpoint Tmot_sp of said electric machine and / or a command setpoint ECCsp of said kinematic chain.

3. Method according to any of the preceding claims, wherein said torque setpoint of said propulsion system is a filtered torque setpoint TPTflt_sp.

4. Method according to Claim 3, wherein a torque setpoint Teng_sp of said combustion engine and / or a torque setpoint Tmot_sp of said electric machine are determined by means of said filtered torque setpoint TPTflt_sp of the propulsion system and steps a) to c) are repeated to determine a command setpoint ECCsp of said kinematic chain by means of said unfiltered torque setpoint TPTsp, and said commands are applied to said hybrid propulsion system.

5. Method according to any of the preceding claims, wherein said discretization takes into account the state of charge of the system for storing electric energy and / or the speed of the propulsion system.

6. Method according to any of the preceding claims, wherein said consumption mf of said combustion engine is obtained by means of a map.

7. Method according to any of the preceding claims, wherein said polluting emissions mNOxEO output from said combustion engine are determined by a model or a map.

8. Method according to any of the preceding claims, wherein the temperature of said after-treatment system is estimated by means of a formula of the following type: T AT t = T AT t − Δ t + Δ t × h 1 t + h 2 t I with: h 1 t = k 1 × T 0 − T AT t − Δ t h 2 t = k 2 × T AT QS u 1 t − Δ t , u 2 t − Δ t − T AT t − Δ t T AT QS u 1 t − Δ t , u 2 t − Δ t the temperature measured in the after-treatment in the steady state, Δt a time increment, k1 the equivalent thermal resistance of exchanges with the exterior, k2 the equivalent thermal resistance of exchanges with the exhaust gases, and I the thermal inertia of the after-treatment system.

9. Method according to any of the preceding claims, wherein the minimization is carried out by means of Pontryagin's minimum principle.

10. Computer program product that is downloadable from a communication network and / or recorded on a medium that is readable by computer and / or executable by a controller, comprising program code instructions for implementing the method according to any of the preceding claims, when said program is executed on a controller of a hybrid propulsion system.

11. Hybrid propulsion system of a vehicle comprising at least one electric machine, at least one combustion engine, at least one system for storing electric energy powering said electric machine and at least one system for after-treatment of polluting emissions of said combustion engine, and a controller configured to implement the control method according to any of Claims 1 to 9.

12. Vehicle, in particular a motor vehicle, comprising a hybrid propulsion system according to Claim 11.